Type 1 diabetes mellitus (T1DM) is an autoimmune disease resulting from the T cell mediated destruction of insulin-producing beta cells located in the pancreas. Current treatment, which includes insulin replacement by injection, frequent blood glucose monitoring, and dietary/exercise discipline, can prevent death from hormonal insufficiency, but is not curative and does not completely prevent the long-term complications including nerve damage, and vascular damage to both large and small blood vessels. [unreadable] [unreadable] In previous few years, we developed a transgenic mouse to express the T cell costimulatory receptor CD80 on its pancreatic insulin producing beta cells (under the control of the rat insulin promoter abbreviated RIP) and demonstrated the mouse's extraordinary sensitivity to autoantigen induced immune mediated beta cell destruction, and thus to diabetes. We refer to the diabetes induced in these RIP-CD80 mice as experimental autoimmune diabetes (EAD). Using this EAD model, we've found that relatively weak anti-beta cell autoimmune responses can cause chronic progressive and eventually complete beta cell destruction resulting in symptomatic and irreversible disease. The slow but inexorable process is highly reminiscent of the beta cell destruction leading to clinical T1DM in man, typically months to years of anti-beta cell immune activity precedes sufficient beta cell killing for the blood sugars to rise. Most other autoimmune diabetes studies involve the non-obese diabetic (NOD) mouse, which develops spontaneous diabetes, or virus-induced diabetes models. Only the RIP-CD80 transgenic mouse diabetes model is characterized by an experimental genetic susceptibility trait (in the EAD model the trait is the CD80-transgene) rendering the mouse susceptible to autoantigen-specific T lymphocyte sensitization as is thought to be present in T1DM patients.[unreadable] [unreadable] We have reported that immunizing with either experimentally-introduced autoantigen (e.g. pancreatic beta cell-expressed viral glycoprotein) or endogenous beta cell autoantigen (e.g. insulin) could lead to diabetes in RIP-CD80 mice. Further, most studies now support that the normal individual's T cell repertoire contains potentially autoreactive but quiescent T cells. We concluded that pancreatic beta cell likely contains many autoantigens and that effective control mechanisms must exist to prevent autoimmune responses in healthy individuals. Current immunological dogma suggests that naove CD8 cytotoxic T cell (CTL) precursors respond to strong antigen stimulation in a characteristic fashion by: (i) proliferating and releasing inflammatory cytokines, (ii) differentiating into CTLs and, (iii) down modulating certain surface interaction molecules to allow the CTLs to leave secondary lymphoid organs while increasing other receptors, such as integrins, to promote entry into peripheral tissues, and (iv) changing chemokine signals necessary to facilitate the relocation of activated CTL from lymph nodes (LN) to inflamed peripheral tissues. [unreadable] [unreadable] The EAD model has allowed us to study the response of beta cell-specific CD8 T cells to cognate antigen presented by either professional antigen-presenting cells (APC) like mature dendritic cells DC), or by non-professional (np) APCs like fibroblast-like cell lines (FCL). While DC-stimulated T cells produced the expected effector CTL phenotype described above, FCL-activated T cells were quite different. Relative to DC activated CTLs, the FCL-activated CTLs proliferated less, released equivalent proinflammatory cytokines, and surprisingly displayed increased cytolytic function. Moreover, FCL-stimulated T cells largely failed to switch their homing receptors predicting poor migration from the lymph nodes to the periphery. Most strikingly, however, FCL-stimulated but not DC-activated CTL expressed many of the features associated with memory CTLs; both multiple memory cell surface marker expressioon and predominant homing into secondary lymphoid organs upon adoptive transfer into naive mice. Importantly, while every CTL response gives rise to a small population of long-lived memory CTL, the mechanism that decides the fate of an individual CTL to become a memory cell remains largely unknown. Finally, the FCL-stimulated CTL induced diabetes by a slow, chronic process, suggesting that these central memory CTL might also be involved in driving human T1DM. [unreadable] [unreadable] Research goals pursued in FY 2007: [unreadable] [unreadable] 1. We developed a potent cell immunization strategy using mature dendritic cells loaded with antigenic peptides (derived from beta cell-made proteins), and established a clear antigen-dose diabetes-response relationship for two experimental autoantigen determinants (peptides), encoded by the transgenic LCMV-GP specifically expressed by beta cells. We hypothesize (and we are testing) that antigenic peptides representing genuine beta cell target epitopes will similarly mediate beta cell killing and diabetes development in this test system.[unreadable] [unreadable] 2. So far we have identified one beta-cell derived peptide (from a serine/threonine kinase called dystrophia myotonica kinase (DMK) aa143-151) capable of sensitizing immune cells to destroy beta cells. This antigenic peptide has previously been reported to be able to mediate diabetes in another mouse diabetes model, the non-obese diabetic (NOD) mouse. Other H2b-restricted antigenic peptides are being tested, but none of several other published autoantigenic peptide candidates have demonstrated the ability to activate a genuine anti-beta cell response in our test system.[unreadable] [unreadable] 3. While only a small number of the many potential beta cell antigenic epitopes have been characterized, many beta cell specific proteins have been linked with the autoimmune proteins leading to T1DM. Using a retroviral targeting approach, we are developing and refining beta cell protein expression in short-lived dendritic cells. This sytem allows exposing the immune system to beta cell candidate protein antigens presented in a highly immunogenic fashion by the transduced dendritic cells, and can be precisely controlled (to quantitate beta cell protein expression), so that dose response relationships between candidate autoantigen peptides can be precisely compared. [unreadable] [unreadable] 4. In parallel, we are developing comparable approaches in mice "humanized" to express a chimeric human HLA-A2 linked to the murine H2K(b) C-terminal sequences. This membrane molecule has been shown to function as a MHC classs I molecule in mice yet present human (HLA-A2-restricted) epitopes to mouse T cells. Autoantigenic peptides identified using these mice could be used to test autoreactive T cells in HLA-A2 positive subjects with T1DM (about 50 percent of the population).[unreadable] [unreadable] 5. Immune cells in draining lymph nodes are continuously exposed to autoantigen from peripheral organ cells. We have observed that beta cell specific T cell activation depends on the cognate antigen but also on the context of T cell activation. Sub-optimally activated T cells (using either altered peptide ligands (APL) or non-professional (immature) APC) display both bridled effector lymphocyte functions (killing, cytokine secretion) and migration. These defects resulted in markedly reduced diabetes incidence in our EAD model. Current efforts are focused on strategies to induce migration-defective T cells that would be maintained in the draining lymph nodes by presense of autoantigen and there work to down-modulate the anti-beta cell immune response.